Ok, it dependes on what do you want to get! I used that sequence just to check if I have fungus DNA if it is positive for PCR I use a specific sequence.

I performed phylogeny among within genus level of Buthus and Mesobuthus scorpions. The level of genetic distance is maybe 1-5%. I can imagine that depending on which species one works there might be quite some difference in this region. Generally, one may say there should not be any selection in this region, however, based on the level of variance I found so far I would claim this region is under selective constraint as well.

So, within species I did not find many alleles in scorpions only if I compared different species.

In my opinion, the dfference of the horizontal comparison is more significant than the difference of vertical comparison. In other words, the parents has not significant differences with their generations(just as you said, use conventional breeding) in ITS sequences, but there are significant differences between the different species, even the subspecies. Because non-function sequences in the DNA have less pressure from evolution than function sequences, where the ITS(internal transcribed spacer) belongs to the non-function sequences. SO in the analysis of ITS sequences mostly be used to distinguished the different species with different characters(1) or subspecies or identify the fungi with special feature (2). And if the parents and their generations have great differences in ITS sequences, how it can be used for identifying?

By the way, nuclear fusion is belongs to mutation, and the consequences may include the change of ITS sequences. But the conventional breeding, even the tissue culture will not cause the high frequency mutation of ITS sequences.

(1) Clark, Curtis. 1995. Reconciling phenotypic and ITS sequence phylogenies in the Encelia alliance (Asteraceae: Heliantheae) to study character evolution. Amer. J. Bot. 82(6), Supplement, p. 120.

(2) Chen,Y-C, J. D. Eisner, M. M. Kattar, S. L. Rassoulian-Barrett, K. Lafe, A. P. Limaye, and B. T. Cookson (2001). "Polymorphic Internal Transcribed Spacer Region 1 DNA Sequences Identify Medically Important Yeasts". J. Clin. Microbiol. 39 (11): 4042–4051. doi:10.1128/JCM.39.11.4042-4051.2001. PMC 88485. PMID 11682528.

I recommend you no to use ITS for vertical comparison (as you commented), you must find a high variable gene to do your pourpose. Probably mitochondrial restriction patterns or delta (transposable elements) pattern can help you. Good luck.

It's all depend on the question you want to adress! Basically I agree with David. from my own experience with yeast ITS present little (but some) variation at the intraspecific level. mDNA RFLP is an efficient technique and AFLP should give better results but requires more work . Good luck.

ITS is good for horizontal comparison and we have done it for many pathogenic species. It may not give much information in vertical comparison as the changes expected are almost minimal or even nil

What exactly is your aim in doing the inter-specific, horizontal comparison? Do you like to investigate familiar structures? I think, for this purpose classical markers, like microsattelites, are a good choice. However, it would seem very interesting to study the per-generation mutation dynamics of rDNA. This is complicated to do since the rDNA locus comes in the order of 100 copies per genome (in animals).